A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in Geotechnical and Earth Resources Engineering

Thesis (M. Sc.) - Asian Institute of Technology, 2025

Driven piles serve as a vital component in transferring structural loads to deeper, more stable soil strata. Predicting their axial capacity accurately is essential for safe and efficient foundation design. This research investigates the combined influence of energy transfer efficiency and time effect (set-up) behavior on the axial capacity of driven piles at both End-of-Driving (EOD) and Start-of-Restrike (SOR) stages. Existing prediction models often overlook the combined effects of these two (2) factors, leading to underestimate or overestimate the actual pile capacity, resulting in wide variation between predicted and actual axial pile capacity.Field past data from dynamic pile testing and driving records were analyzed to quantify energy losses during installation and capacity gains from soil recovery over time. Results showed that average hammer energy transfer efficiency was 57.85% at EOD and 40.32% at SOR. These values varied notably with ram weight, pile type, and soil condition, with steel piles and heavier hammers achieving the highest efficiencies. Time effect (soil set-up) analysis revealed a mean set-up factor of 1.612, indicating a 61.2% increase in pile capacity due to rest periods, with shaft resistance identified as the dominant contributor to this gain.The capacity prediction analysis demonstrated that the Danish formula offered the most consistent results at EOD, with a low coefficient of variation (COV=0.22) and a reliability index of Ý = 0.88. For the SOR condition, the Janbu formula with integrated set-up correction yielded the highest accuracy, achieving the lowest root mean square error (RMSE= 0.237) among all models evaluated. These findings affirm that predictive reliability improves significantly when both energy transfer and time-dependent (set up) behavior are incorporated into the analysis.The proposed model enhances the accuracy of driven pile capacity estimation by unifying energy efficiency and set-up effects into a model applicable to both immediate and long-term conditions. This approach provides engineers with a more reliable tool for foundation assessment, supporting safer design decisions and cost-effective construction practices in varied geotechnical settings.